Reconstruction of diffusion using spectral data from time series (Q864092)

A numerical technique for the reconstruction of drift and diffusion coefficients of stochastic differential equations from real data such as time series is presented. The coefficients are found by minimizing an object function which measures the difference between the eigenspectrum for the infinitesimal generator of the related diffusion process and a reference eigenspectrum from time series by constructing a disrete time Markov chain. By this procedure, the minimization of the object function is turned into a quadratic programming problem on a convex domain for which well-established solution methods exist.